Homelessness and Residential Mobility: Risks for Later Academic Achievement

• In Event: Poster Session 02
  In Poster Session: PS 02 Topic - Family Context & Processes

Abstract

Research suggests that both homelessness (Rafferty & Rollins, 1989) and residential mobility (Coley & Kull, 2016) are associated with lower child academic achievement. Most studies that examine these variables in the same model tend to combine them into one construct (e.g., Cutuli et al., 2013), rather than exploring how they may differentially affect achievement. Examining them separately could clarify whether homelessness is associated with poor academic outcomes above and beyond other forms of mobility (e.g., school mobility; Fantuzzo et al., 2012). Academic resources and parent involvement have been linked to children’s achievement (e.g. Jeynes, 2007, Tucker-Drob & Harden, 2012). We sought to examine individual effects of homelessness and residential mobility on academic achievement and investigate whether access to reading, developmental stimulation, and parent school involvement would moderate risk.

We utilized longitudinal data of 4,898 families from the Fragile Families and Child Wellbeing Study (FFCWS). Counts of both homeless episodes and residential mobility were derived from mother’s reports of homelessness and moves collected at ages one, three, and five. At age five, mothers reported on child’s access to reading and developmental stimulation. Parental school involvement was measured at age five utilizing maternal-reported frequency of attending school meetings. Children’s academic achievement at age nine was constructed by averaging standardized scores for two subtests of the Woodcock-Johnson (WJ; i.e., passage comprehension, applied problems) and the Peabody Picture Vocabulary Test (PPVT). We hypothesized that more homeless episodes and more residential moves from birth to age five would each predict lower academic achievement at age nine. We hypothesized that academic resources and involvement at age five would each interact with risk to attenuate these associations.

The path analysis model displayed good fit, Chi-square (df = 51) = 199.60, p < .001, RMSEA = 0.024, CFI = 1.00, TLI = .98, and SRMR = .02 (Table 1; Figure 1). In line with hypotheses, more homeless episodes predicted lower academic achievement at age nine (b = -0.06, p = .001); however, contrary to expectations, more residential moves predicted higher achievement (b = 0.04, p = 0.03). Developmental stimulation at age five was associated with higher academic achievement at age nine (b = 0.09, p < .001). Greater access to reading predicted higher academic achievement at a trend level (b = 0.05, p = .05) while parental school involvement was not associated with academic achievement. Access to reading, developmental stimulation, and parental school involvement did not moderate these associations.

Findings suggest that homelessness predicts lower academic achievement beyond residential mobility, academic resources/involvement, and a number of associated sociodemographic risks. Results suggest that homelessness may account for much of the negative impact of residential mobility such that other residential moves actually predict better achievement. This may be due to the potential higher incidence of upward trajectory moves, and their associated long-term protective factors, after controlling for homeless episodes. Future research should avoid grouping both populations and instead seek to examine the unique characteristics of each. Further findings and implications will be discussed.

Authors

• Joanna N. Keane, Northern Illinois University
  Presenting Author

• Maria Y Abdul-Masih, Ohio State University
  Non-Presenting Author

• Elizabeth C. Shelleby, Northern Illinois University
  Non-Presenting Author

• Samantha Awada, Northern Illinois University
  Non-Presenting Author

• Janette E. Herbers, Villanova University
  Non-Presenting Author

• Brianna I. Banks, Northern Illinois University
  Non-Presenting Author